Those Who Can Do, Those Who Can't, Get Certified

As I think about the question of whether data professionals should pursue data mining or other certifications, my thoughts first head back in time.

In the old days, if you went to get a proper Computer Science degree, you'd have to take a course called "Theory of Computation." This was a generally pretty useless course involving tons of discrete math and lots of odd theory. To save the pain, I'll boil the class down to this takeaway: "Given a complete specification of the input, computers are black boxes that take that input and produce output according to some mathematical equation." The output is always nothing more than some function of the input. (I hope you're thinking "Well duh!" -- if you're not, let's discuss it in the comments.)

Colloquially, we used to call that class "Job Security 101" (partly to make it less painful). It's important to know what types of things computers can in theory do well -- because it's pretty likely that if they can do them in theory, they'll eventually do them in reality. Applied to data analytics, this is suddenly a discussion all about certifications.

To get a certification, you must pass a test. The test will ask you questions of the form: “Here is a question, what is the right response?” The question is the input, and the astute will see the parallel -- this is exactly what a computer does best. Inherent in these sorts of tests is a fundamental problem: They’re tests that a computer can solve (in theory).

At the pace of modern computing technology progression, are you sure that’s the business you want to get into? How long do you think you can compete? If a computer can do it in theory, then it'll soon replace you. Is trying to stay ahead of that curve the sort of rat race you want to be in for the rest of your career? This is not a plan for success, I say. Don’t build the foundation of your career on these stones -- especially if your career is in the increasingly important area of data analytics. If, in theory, a computer could ace an online data-mining exam, then consider making sure this isn’t your foundation.

Have you ever observed that a certification in 1998 isn’t quite the same as the same certification in 2008? While I sounded theoretical above, I’m really talking about hard practicalities. Computer advances really are the reason that certifications get outdated. This isn’t just theory talk here. What you certify for this year will be outdated next year because the computer will be doing more of your job.

To paraphrase Albert Einstein: Education is what’s left over when you’ve forgotten everything you’ve been taught. I am absolutely supportive of getting more education -- do it. But make sure that’s what you’re getting, as opposed to being required to give the right answer to a given set of inputs.

And consider two additional points.

In some professions, it’s not what you know, but who you know. I've experienced this personally -- every job I’ve ever gotten has been through someone I know. But my own experience also has been that success has come from being able to think on my feet, read people, communicate well, come up with creative solutions, and deal well in situations where there are no “great” solutions. These are the skills that make success in a job, but they are not tested in these certification exams. Sociologist William Bruce Cameron apparently once said, “not everything that’s countable counts, and not everything that counts is countable” -- and in my experience, most of what makes for success in a job is in the “uncountable” category.

And one last quote: “There comes a time in life when you need to stop sharpening your arrows and start shooting with them” (rumored to have been uttered by an old Indian, but the Internet doesn’t find this citation). When I participated in hiring committees at Google, lots of certifications was generally considered a negative signal. Usually this came from a feeling a mal-prioritized time -- is there nothing better the job candidate could have done with his or her time? Why not accomplish something? You learn so much more by actually getting dirty doing things than you do studying for a test -- show us where you got dirty and what you learned and what you contributed.

And I think that really is the bottom line: Go and do something valuable instead of studying for a certification exam. What you’ll learn is, believe it or not, the solution to the job security problem.

Share your thoughts in the comments!

Take the quick poll on certifications, at right, too!

Point / Counterpoint, Founder, Better Place Consulting

Counterpoint: Train for Data Visualization Skills

Chances are you've already got good data visualization experts on staff, even if you don't know it yet.

Customer-Centric Banking Analytics Scares Me

Banks have tons of customer data at their disposal; unfortunately not all will use it scrupulously.

Certification counts for something
  • 12/20/2012 9:01:11 AM


Masam writes

Today is a world of professionals and everyone prefer specialists


I agree that in many fields, certification in a speciality, particularly an especially "hot" speciality, can improve your marketability and career prospects.  Data mining is probably one of those cases.


Re: Some people just get it
  • 12/20/2012 7:15:00 AM

@Scott and @philsimon

Today is a world of professionals and everyone prefer specialists - Specialist in medicine, defence, education, IT, HR and management. Everyone wants to pay for reliable and trustworthy things.


Certifications are really beneficial if and only if you "Do it once and do it right" full stop and afterwards every next iteration is just to groom same things using experience of known specialists – who are offering certifications. You cannot change your brain like computer but you can upgrade with new and relevant material.


Persons who have concerns about certifications are possibly those who due to any reason could not do it right first time themselves.  Either they have tried to pass their certification using some nasty ways or at least they are well aware of known cheats and backdoor which have created mistrust in their minds.


On first day of carrier, not everyone is lucky from ancestors or have good relationships in the industry of their choice. In that situation, a certification/degree is necessary to show some commitment to their carrier plans and most improtantly bring one on the interview table and after that it's all up to the interviewer - "Only smart people can hire smart employee". If the interviewer is very clear about his/her own carrier path and they also knows how the energies and capabilities of new person can best fit in the overall vision of their organization, only then they will choose the right candidate for their team.



Re: Some people just get it
  • 12/16/2012 8:53:17 PM

One indication of the power of certification is the talk now surrounding online education platforms like Coursera. The argument is that they will not become relevant (and profitable) until they offer certifications through their courses --- instead of just knowledge for knowledge's sake.

Re: Some people just get it
  • 12/10/2012 11:41:37 AM

Agreed. Give me a non-certified expert over a certified newbie any day of the week and twice on Sunday. I was certified in Lawson Software in 2000 after a mere 2.5 months. Eigh years later, I was finally 'recertified' and my increase in knowledge between those two periods was exponential. Official certification doesn't mean much by itself.

The CEE Triad
  • 12/8/2012 8:56:18 AM

It is very hard to take sides in this issue. I can see valid arguments on both sides. Personally, on the one side I feel certifications are quite expensive, the material learned and certified on gets stale very quickly, and I feel very strongly that the certifiers (ISC2, Cisco, etc.) should provide a free second try for taking the exam should one fail it the first time. Six hundred dollars a pop can leave a serious dent in one's finances, not counting the price of the books, study material, and classes. Certification classes demonstrate that one can memorize the material and pass the test, but what happens when the successful candidate is thrown into the real world and they have to prove what they know?

On the other hand, certification does not take the place of education and experience... it ADDS to it. The study material, lectures, boot camps, etc. are a valuable asset to continuing education and experience. I do concede that the certification exposes the candidate to a wealth of information that will be of significant value to an engineer, administrator, etc. in the performance of their duties.

I personally feel that the CEE Triad (Certification, Education, and Experience) make up the necessary ingredients of making and shaping a properly educated and experienced professional.

Re: Some people just get it
  • 11/29/2012 2:04:01 PM

While I agree that being over certified is useless as well as old certifications, I do think they can be useful in certain areas such as broadening your knowledge base or acquiring a skill you weren't trained in. It would be great if we got this in the workplace but not all employers encourage on the job training. I had an individual that worked with me and they weren't a good fit in their current position, I encouraged them to seek a certification which opened an entire new job path for them that they are doing fabulously in today. It can work!

Re: Some people just get it
  • 11/28/2012 5:58:54 PM

@mnorth and @Beth, ere's my personal reference point on that last point:

My MS advisor didn't like it when we called him "Dr" - he said that designation is important only to those who don't have it.  I've seen his point, but somewhat disagree - it's also very important to those who have paid the price, and in my case the biggest price was paid by my wife while I was earning it.  It means a lot to her, probably more than it does to me.

In earning my PhD, I think the three most important things I learned (unfortunately, because none of these are "computer science" or anything like that) are these: 

1) It doesn't always take what we think to earn one - sometimes much less, sometimes much more.  Often the general expectation of someone with one is reasonable, but some people paid a whole lot more, and some people paid very little.

2) Having one opens doors

3) Having one closes doors.

The doors it closes are not the same ones it openes, and which it opens often surprise me, and which is closes often surprises me too.


In the end, it was definately the right thing for me to do, but the effect has been a very different flavor than I thought it would be.

Re: Some people just get it
  • 11/28/2012 4:49:07 PM

Re: cheating: Agreed Beth.  Not every college graduate cheated their way through (I didn't), and finding the ones who did by looking at resumes it very difficult.  Ultimately, you want to hire the person who can do the job for you, and do it well.  Resumes, college degrees, certifications, etc. have all be born out of the need to determine from paper which applicants can, and which can't.  They're imperfect instruments, I'll grant you that, but we use them because they're effective, or at least we've convinced ourselves they are.  

I recently saw an article, that I can't find now, about highly successful people who never finished college (and who I'd assume, don't have certs either).  The point of the article is all about what you can do, not about the size of your honorary portfolio.  I think they'd side with Noreen -- I'll take the one who can do every time over the one who's studied a lot and collected a bunch of paper.

Is it fair though to assume that a college degree or a certification doesn't indicate an ability to do?  Some certifications cannot be earned without a significant amount of demonstration of the ability to do.

Re: Some people just get it
  • 11/28/2012 4:41:56 PM

@PC: I can agree to an extent.  The problem was that the job market my dad was in was flooded at that time because of defense spending cuts.  The guys my dad knew were either a) laid off with him (and thus competing with him in the job market), or b) faced with many qualified applicants who were easily separated by whether or not they had paper credentials.  My dad's application kept going in the 'have-not' pile until eventually, someone he knew was able to hire him without getting criticized or hassled by some higher-up for passing over more credentialed applicants.  The 'know someone' approach worked for my dad, but it took longer than if he'd had the papers.

Also, the 'know someone' approach has not been my experience.  I have worked for four employers thus far in my career, and in all four instances, I obtained my job without knowing anyone on the inside.  Perhaps I'm an exception to the rule, or perhaps I don't play the game right, but thus far my credentials on paper have had to speak for themselves. I feel very fortunate that that approach has worked for me thus far. 

Re: Some people just get it
  • 11/28/2012 4:26:36 PM

Well sure, especially when you put it that way. But picking out the cheaters, from a resume alone, is tough. 


Page 1 / 2   >   >>